Conventionally, an operation apparatus for an in-vehicle device such as a car navigation device has been proposed. For example, such an operation apparatus obtains a photographic image of a hand of an operator with a camera, extracts a finger image from the obtained photographic image, and superimposes the extracted finger image on a graphical user interface (GUI) input screen such as a navigation screen of an in-vehicle device. For example, U.S. Pat. No. 6,407,733 (JP-A-2000-335330) discloses a device configured to obtain a photographic image of a hand of an occupant, which manipulates a touch panel near a seat of the occupant, using a camera mounted on a ceiling of a vehicle body. The device is further configured to indicate the obtained photographic image of the hand and a touch panel on a liquid, crystal display, which is opposed to the occupant. JP-A-2001-216069 discloses an operation input device configured to assign different operation contents according to the distance between a screen of a display device and a user's finger or the like.
In the conventional operating systems, the image information such as the image of a hand obtained with a camera is used only for superpose-indication of an outline image of the hand for indicating the operation position on the screen. Thus, the image information may not have been effectively used as input information. In particular, during a driving operation of a vehicle, a driver cannot gaze at the screen of the operation apparatus. Accordingly, the driver tends to cause an erroneous operation. In addition, it is hard to perform a complicated operation during a driving operation of a vehicle. The operation apparatus of U.S. Pat. No. 6,407,733 uses two input information including the photographic image information on a hand and touch detection information. However, the two input information may not be effectively utilized for enhancing an operability.